Brian Friel (E-mail this author)
© National Journal Group, Inc.

In the fall of 2003, more than 400,000 preschoolers in the federal Head Start program sat down individually with their teachers for an academic assessment. During each 20-minute test, the kids named as many of the letters of the alphabet as they could. They looked at a series of pictures and picked the ones that best represented "swamp," "diving," and other vocabulary words. They tried basic math skills, such as counting to 25. In the spring of 2004, the same children worked through the assessment again. Then the Head Start program and its test contractor tabulated the results to see how much the kids had learned over the school year. It was the largest standardized assessment of preschoolers, ever.

But Head Start officials did not announce the results publicly. The findings were shared with Head Start directors, but many in the education world did not even know that the assessment had taken place. Wade Horn, head of the Health and Human Services agency that oversees Head Start, said the test was a practice round, designed to work out any kinks. "You have to understand my 'first-pancake theory,' " Horn said. "The very first pancake I make, I never serve. The first pancake is a test pancake. Do I have the batter right, the heat correct? I make adjustments and then the next pancakes, those are the ones I serve."

But Sarah Greene, president of the National Head Start Association, said that HHS chose not to share the results of the test because the results make Head Start look good -- and because, she says, the Bush administration wants to dismantle the program. "We're not going to let HHS hide our lamp under a bushel or basket," Greene said. "The assessment provides convincing proof that the program works as intended." At the beginning of February, Greene's group released the results, which showed that Head Start kids made gains in both language and math skills.

The test -- and the controversy over releasing the results -- exemplifies the predicament that Head Start confronts in its 40th year. Designed to help the nation's neediest children prepare socially, emotionally, and academically for school, Head Start is facing unaccustomed scrutiny from the Bush administration and Republican lawmakers as Congress debates legislation reauthorizing the program.

Critics question Head Start's effectiveness in preparing kids for school and its management of funds. "Head Start is not quite like Social Security, but it's got an iconic quality such that people aren't supposed to say anything bad about it," said Chester Finn, president of the Fordham Foundation, a conservative education-policy group. "It's done a lot of kids a lot of good, but it desperately needs modernization and a makeover." Head Start advocates argue that the program is already rigorously reviewed and that it has proven its worth; they see the criticisms as a smoke screen masking attempts to discredit a Great Society program that works. As the 109th Congress starts, the various players are seeking to frame the debate: Is Head Start in need of major reform, or just a little tweaking?

Started in 1965, Head Start provides an array of services to 910,000 low-income toddlers and their families. About 1,700 local Head Start providers -- including nonprofit organizations and school systems -- receive nearly $7 billion in federal grants to offer education, child care, medical and dental care, nutrition, and other services to the poor. The grantees operate 19,200 Head Start centers in all 50 states, the District of Columbia, Indian reservations, Puerto Rico, and other U.S. territories; the program has a total paid staff of 206,000 employees.

Head Start's latest reauthorization has been stalled in Congress for nearly two years. In July 2003, the House passed a reauthorization bill by one vote -- 217-216, with most Republicans voting yes and all Democrats voting no. The controversy centered on a provision that would give more control to states. Since many states in recent years have set up extensive preschool programs, the provision would have let eight states control Head Start dollars and coordinate Head Start with their state programs.

"I happen to still believe it's the right way to go," said Rep. Michael Castle, R-Del., chairman of the House subcommittee in charge of Head Start. Democrats, spurred by the National Head Start Association, viewed the provision as a threat, arguing that it would turn over a successful federal program to states that have less-effective preschool programs. "You're just downgrading the Head Start program with that approach," said Rep. George Miller, D-Calif., ranking member of the House Education and the Workforce Committee. "There were many states that would have gotten the block grant that had no experience in running quality programs at all. You would turn this program over to the states, and they could start bleeding services off."

In the Senate, key Republicans, including Lamar Alexander, R-Tenn., opposed the devolution plan. In October 2003, the Senate Health, Education, Labor, and Pensions Committee, on a unanimous, bipartisan vote, approved a Head Start reauthorization bill that did not include the provision. But with disagreement among Republicans about the devolution plan, the bill never got to the floor.

Instead of devolution, the Senate bill required more-rigorous reviews and audits of Head Start programs. The National Head Start Association opposed that approach, arguing that HHS already rigorously reviews Head Start programs. "We felt that it would be a waste of dollars," the association's Greene said.

The Bush administration, meanwhile, had already begun its own plans for a national test, called the National Reporting System. Horn said that the Administration for Children and Families, the HHS agency that oversees Head Start, reviewed a variety of approaches to the test, including measuring social development. But Horn said he was comfortable with the validity of only a few academic measures, on early math skills, vocabulary, letter recognition, and understanding of spoken English. The testing is consistent with the philosophy embodied in the 2002 No Child Left Behind Act, which tests elementary and middle school students with the goal that all students will be reading and doing math on grade level by 2014.

Yasmine Daniel, director of the early childhood development division at the Children's Defense Fund, says the Head Start test is "developmentally inappropriate," meaning that the measures aren't a proper test of the program's goal of getting children ready for school. Some of the test measures are culturally inappropriate too, she said. For example, an inner-city kid would have trouble identifying pictures that correlate to "diving" or "swamp," she said. In addition, testing academic progress leaves out other important measures of social and physical readiness. "We definitely don't believe we should be testing 4-year-olds," Daniel said.

Nick Zill is the director for the National Reporting System at Westat, the Rockville, Md.-based firm that is HHS's contractor for the testing project. He said the tests were strictly reviewed for quality and have been carefully constructed with 4- and 5-year-olds in mind. (Head Start enrolls some 3-year-olds as well, but no one is suggesting that they be tested.) He disagreed with Daniel's view on cultural appropriateness. "Children are not exposed to astronauts or princesses or dinosaurs, either," Zill said. "It's not the fact that a child does or does not know a particular word that matters. It's whether their vocabulary expands from fall to spring."

In the first-year results, Head Start kids improved their scores in all areas. For example, from fall to spring, the percentage of children who knew fewer than 10 letters fell from 74 percent to 39 percent, while the percentage of kids who knew 17 or more letters rose from 16 percent to 46 percent. Zill said the results are mixed. "There's impressive growth, but at the end of Head Start, they are still behind" children from higher-income families, he said. "We think there is room for improvement."

The National Reporting System seems to mirror findings from other Head Start evaluations: Head Start helps kids move up but doesn't entirely close the gap between poor children and wealthier children. Advocates worry that the Bush administration plans to emphasize the negative side of that equation in an effort to defund the program. "One of the problems for many in Head Start is the way in which the whole discussion about assessment occurred. President Bush said Head Start is not working well, not meeting outcomes," said Ron Herndon, chairman of the National Head Start Association and director of a Head Start program in Portland, Ore. In July 2003, Bush commented about Head Start programs: "They're working OK. We want better than OK in America. We want excellence."

Herndon said, "If someone had come up with a perfect assessment in that atmosphere, there would have been suspicion."

Horn said that the administration will use the test results to provide additional assistance to struggling programs, some of which provide as many anti-poverty services as they do academic services. "The first response, if a program is consistently showing little progress, is training and technical assistance," Horn said. "I can't, for the life of me, understand why anyone would think it's a bad idea to assess whether a program is progressing in critical academic areas."

While Head Start's academic effectiveness will be a topic for debate in Congress this year, House Republicans also plan to focus attention on management of the program. Last year, newspapers reported on Head Start directors making up to $300,000 a year at the same time that some teachers in the program earn only $7 an hour. Rep. Castle and Education and the Workforce Committee Chairman John Boehner, R-Ohio, subsequently asked HHS for more information about the program's salaries, travel budgets, and financial management. Castle said the information he's getting back troubles him. "I'm not entirely satisfied that Health and Human Services is monitoring these programs as well as they should be," Castle said. "There are questions about the management of these programs that need to be looked at."

Horn said the administration is revamping its oversight of Head Start. "There have been far too many instances in which we have found out that local Head Start programs lacked adequate financial controls and oversight," Horn said. But he also said that he will seek more flexibility from Congress, arguing that the law governing the program is overly complex, making oversight difficult and giving the executive branch little power to make changes. "This law is an abomination," Horn said. "Head Start is the most micromanaged program on the face of the Earth. At some point, you have to give your program managers the leeway to manage the program in a way that makes sense."

Castle and Alexander, who now chairs the Senate subcommittee that oversees Head Start, both plan to have hearings on the program's academics and management before they move reauthorization legislation. A key for Alexander is a review of all federal early-childhood programs to see if they can be better coordinated. Lawmakers from both sides of the aisle say they want to start over on reauthorization and work together to improve the Head Start program. But with Republicans running both chambers and convinced that it's time for serious scrutiny, the program's advocates should expect a long year. "Head Start is not something to be put on a pedestal and worshipped," Finn said. "It is something to be fixed and updated to deal with the challenges of the present day."

Testing Head Start

According to the first-ever nationwide standardized test of Head Start toddlers, kids learn a lot in a school year, though critics say not enough.

Percentage of kids who ...

were at or above age level in vocabulary
Fall 2003    20
Spring 2004  31

could identify 17or more letters
Fall 2003    16
Spring 2004  53

could count 15 or more objects
Fall 2003    14
Spring 2004  61

Catching Up

A comparison of Head Start kids to average American kids shows that the program helps close -- but does not completely eliminate -- the academic gap facing children from poor families.

Distance from national average ...

in percentage points